Re: building default not possible?

[John W. Eaton]

On 07/08/2013 10:16 PM, Carnë Draug wrote:
> On 8 July 2013 20:12, Torsten<address@hidden>  wrote:
> > [...] but there are still a bunch of warnings during
> > compilation of __magick_read__.cc:
> >
> > ../../libinterp/dldfcn/__magick_read__.cc:406:1: warning: unused
> > parameter 'args' [-Wunused-parameter]
> > ../../libinterp/dldfcn/__magick_read__.cc:406:1: warning: unused
> > parameter 'nargout' [-Wunused-parameter]
> > ../../libinterp/dldfcn/__magick_read__.cc:874:1: warning: unused
> > parameter 'args' [-Wunused-parameter]
> > ../../libinterp/dldfcn/__magick_read__.cc:1010:1: warning: unused
> > parameter 'args' [-Wunused-parameter]
> > ../../libinterp/dldfcn/__magick_read__.cc:1157:1: warning: unused
> > parameter 'args' [-Wunused-parameter]
> > ../../libinterp/dldfcn/__magick_read__.cc:1215:1: warning: unused
> > parameter 'args' [-Wunused-parameter]
> > ../../libinterp/dldfcn/__magick_read__.cc:378:1: warning: 'void
> > maybe_initialize_magick()' defined but not used [-Wunused-function]
>
> And you didn't had those warnings before? The missing include was my
> fault, but those warnings come from parts I didn't touch. They seem to
> warn about arguments that are used only if you HAVE_MAGICK (and you
> don't). Same for maybe_initialize_magick() which does nothing in your
> case. Should I enclose the whole thing in a ifdef HAVE_MAGICK, rather
> than just the code inside?

No, I think we generally want the functions to be available so they can 
print an error message explaining why they are not functional.

I don't really care about warnings that happen when dependencies we 
expect to be present are missing.  Install the dependencies if you want 
the warnings to go away.

jwe